Wikipedia:Requests for comment/Arbitration Committee/Summary

View by SirFozzie
''I (LaraLove) am posting this on behalf of SirFozzie by e-mail request (sent via Wikipedia interface and confirmed on IRC). This is a straight copy/paste from e-mail, as he wishes to remain on a break.''

Reducing Burnout of ArbCom Members:

Right now, an ArbCom term is 3 years.

I dare you to find ANYONE in this world, who can immerse themselves in every single high drama, night mare conflict for that long and NOT burn out.

And to properly do your job as an ArbCom member, you do have to be hip deep in crap, if you'll forgive the saying.

So, what tends to happen and has been happening, is, the newer folks on the ArbCom do all the public facing work and indepth analysis (Kirill is rather recent, NewYorkBrad before he resigned and now FT2) The rest are generally seen by the community as a rubber stamp. They may discuss in the background, but nothing really happens publicly untill one or two people (usually the same one or two people) write a proposed decision, and 95% of the time, the rest of them come in and sign off on the decision.

This also leads to issues like the Episodes and Characters ArbCom case aftermath, where ANY arbitrator speaking up and clarifying terms of a remedy would have avoided weeks of drama later. No Arbitrator spoke up, and the issue at hand devolved, weeks later.

Or the Cla68/SV/FM/JzG monster case, where there was an admission from a sitting arbitrator that no one had really taken a look at the case, weeks into the case, because everyone was doing their own thing with the end of the school year, vacations, etcetera. At any one time, a quarter to a half of the Arbitration Committee is "inactive", and not doing the job they were elected for.

Perhaps reducing ArbCom terms would allow Arbitrators to devote more "energy"/Research into each case, with less risk of burnout.

I would suggest keeping the three staggered arbitrator groups, but reducing an ArbCom term to 12 months (that means elections every 4 months, which is frequent, but not an everyday situation) or 18 months (elections twice a year)

Once you finish a term as an ArbCom member, you have to sit out one election. (To recharge the batteries, reduce chance of burnout)

Various suggestions by Neil (Part 1)
Moreschi, above, has identified one of the root problems. Arbcom has been unable to scale itself to the community, I believe due to the increase in size of the community, but also due to its maturation; it's not "the same as it was three years ago but twice as big". Things have changed, a lot, and Arbcom hasn't.

Arbcom is becoming increasingly more incompatible with the community it is supposed to serve. We need some form of higher dispute resolution beyond RFC, so let's operate on the assumption we need some form of Arbitration Committee, and there's plenty of evidence that the current one is not fit for purpose in the present day.

So, how can Arbcom change? I have borrowed an idea or two from above and fiddled with them, and added my own. I have split them into subheadings, as of course some ideas may be good, some may be bad. For clarity, and because some follow on from one another, I have put my suggestions together, rather than spatter them throughout this RFC.


 * I. Expansion

'''The three current Tranches are expanded from 5 members to 7. An additional Tranche Delta is created.'''

My thinking: The Arbcom is too small. More members are needed - presently we have just eleven active arbitrators. We need more. I suggest the simplest and most efficient way to address this is to expand the arbitrator Tranches, and add a further one.

I suggest an election be held as soon as is realistically possible, with each Tranche expanded from five Arbitrators to seven, and a further Tranche appointed (Tranche Delta). This would require 14 new arb appointeeships; 2 each for Tranches Gamma and Beta, 3 for Tranche Alpha (the extra one to replace Newyorkbrad) and 7 for the new Tranche Delta. The simplest way would be to appoint the 14 candidates with the highest percentage of votes.

This will give us a maximum of 28 Arbitrators rather than 15. This would still not be too unwieldy. If this still proves insufficient, we can always add a further Tranche.

I am aware the Arbcom has stated on numerous occasions that they do not need more Arbitrators; this ought not to be their decision to make, and the preponderance of evidence is that despite their protests, too many cases take too long to resolve, and receive insufficient Arbitrator scrutiny.


 * II. Timescales

'''Arbitration Committee terms are reduced from three years to two years. Elections will be held every six months on a rotating basis.'''

My thinking: Three years is too long. Shorter terms are needed. Why is three years too long? I believe there are two main reasons. Firstly, fatigue of Arbitrators - easily observed by noting that the newest arbitrators are usually the most active. This would be addressed by both reducing the time served, and by increasing the numbers as detailed above.

Secondly, Arbitrators must have the faith of the community. While it is necessary for Arbitrators to make on occasion unpopular decisions (as what the community wants is not necessarily the sole concern in making decisions), the faith of the community to make those decisions must be maintained.

I suggest an Arbcom term be brought down from three years to two years. This would dovetail neatly with the expansion from 3 Tranches to 4. We should not presume an Arbitrator is incapable of serving longer than this; however, they will have to convince the community they are capable of serving a further two years.

This will necessitate an Arbitration Committee election every 6 months, rather than every 12 months. This is a bearable burden on the community. It would also enable an Arbitrator to, if they wished, to take six months off (a reasonable break) and return to arbitration six months later.

View by User:Barberio
The Arbitration Committee are not entitled to create binding rulings that institute new Wikipedia processes or policy, or substantial alter current Wikipedia processes or policy except by striking sections incompatible with Wikipedia Foundation requirements. They are not a legislature.
 * No New Policy

The Arbitration Policy clearly outlines the form that an Arbitration Case takes. It does not allow for the Arbitration Commity to convene in secret to determine a case based on proceedings conducted entirely in private. While they may accept confidential evidence, proceedings must be as open as they can be.
 * No Secret Trials

No individual, or plurality of Arbitrators should make announcements or declarations on behalf of the entire committee, they are to only be made by the plurality or the individual making them. Dissenting Arbitrators should be given an opportunity to voice their dissents to this action in public if they wish.
 * No Individual Or Plurality Acts To Be Labelled As 'Arbitration Committee Action'

View by Nsk92

 * No secret votes

While it may be necessary for a portion of ArbCom discussions to occur in private (e.g. where privacy considerations and rela-life identities of some parties are concerned), the actual votes on all the ArbCom decisions by the ArbCom members must always be made in public, with the signatures of the arbitrators attached and available for verification through the history log. This way we will know exactly who voted and how, and there will be no room for speculation, error, rumor, misunderstanding, etc. I can see no plausible scenario under which an exception to this principle might concievably be warranted.

View by Durova (2)
Arbitration is the final step in dispute resolution. Except for emergency situations such as wheel wars, or appeals of community sanctions and other circumstances where dispute resolution is impractical, there needs to be prior formal dispute resolution. These options are:


 * Third opinion
 * Requests for comment
 * Wikiquette alerts
 * Mediation Cabal
 * Mediation Committee

These parameters have not been observed consistently. Formal dispute resolution should remain open long enough to have a reasonable chance of success before a request for arbitration is attempted. WP:AN and WP:ANI are not formal dispute resolution; informal mediation is not formal dispute resolution; block warnings are not formal dispute resolution; an e-mail exchange with an arbitrator is not formal dispute resolution. All of these informal methods have been offered and accepted at various times during RFAR. They should not be acceptable and, if proposed, should be removed by the clerks.

The reasons for going by the book are twofold. The exploitation of non-formal dispute resolution as a pretext for arbitration drags unwilling bystanders into cases and fosters drama. Arbitration is slow, time-consuming, and stressful. Editors deserve to interact with each other with reasonable expectations about when an arbitration case may be in the offing and when it is not.

View by Celarnor (2)
I see the need for ArbCom as a binding form of dispute resolution, especially for problems involving admins that can't be handled by a community of editors without the tools to enforce their collective judgement. They have been useful in providing sanctions on editors and admins alike who have gone astray of the rules. However, I do not believe that ArbCom is within its powers to create "remedies" that aren't in the form of direct sanctions on editors and admins listed as being involved in a given case. What they have been doing of late goes well beyond that (The 'Sourcing Adjudication Board' and the 'BLP Special Enforcement' measures come to mind prominently); changing existing policy and creating new procedures is not part of the dispute resolution process; that is part of the governing process, and that is a responsibility that belongs with the community, decided by consensus. Not the consensus of a few elite editors, but the consensus of the project as a whole.

Various suggestions by Neil (part 2)

 * VII. Communication

My thinking: An obvious suggestion, based on recent events, Jimbo's commentary, and common sense. A committee that acts wholly in secret is, by definition, a cabal.


 * IX. Communication

My thinking: Arbcom must recognise that not all Wikipedia editors are familiar or comfortable with jargon such as tu quoque, ''nem. con, sine qua non or ab initio'', particularly when clear English alternatives exist. Arbcom decisions should not have to be read with a dictionary to hand. Recognise that many Wikipedians on en-Wiki may be younger editors, may have poor comprehension skills, or may have a first language other than English. Abbreviations should similarly not be used (and if they must be, wikilink them). Using jargon is unnecessary; it may save the writer a few seconds, but wastes the reader's time far more (and I would trust all writers hope more than one person reads what they have written).
 * Renamed from "No jargon" to "Communication". Neıl 龱  12:23, 30 June 2008 (UTC)

Various suggestions by Neil (part 3)

 * IV. Redress

My thinking: Based on recent events, it is clear that there is no current way for the community to redress the behaviour of an Arbitrator other than voting them out up to three years later. This is not sufficient.


 * Investigation into the conduct of the Arbitration Committee prior to this RfC.
 * (previously titled VI. Transparency)

My thinking: Many users consider some form of investigation into the recent goings on vital in order to restore faith in the capability of Arbcom to act as the final level of dispute resolution on the English Wikipedia. This cannot be done by the current Arbitration Committee - doubts would always remain. Einstein said, "We cannot solve our problems with the same thinking we used when we created them".

Third view by Shoemaker&#39;s Holiday (talk)
The arbitration committee has several times this year acted in ways that prevented or highly prejudiced a user who was not yet able to give evidence. The OrangeMarlin case came with a statement by FT2 that they did not feel like going through the effort to read the user's evidence:. In the Durova and MatthewHoffman cases, highly prejudicial voting began against the accused before the accused - then administrators in good standing - were able to provide evidence.

View by User:Filll on delays and early findings
Sometimes Arbcomm can act extremely precipitously, even violating their own rules, as they did in the Durova and MatthewHoffman cases when they started voting before there was any evidence. In other cases, Arbcomm is unreasonably dilatory and slow, and it creates mega cases that drag on for months. A lot of drama could be saved if Arbcomm made some clear announcements at the start of a case, like "We take the policy on coercion seriously", as a preliminary Finding of Fact (drawing an example from a current Arbcomm case that is languishing). The difficulty arises when there is clear evidence of policy violations, or that potential policy violations are involved, if Arbcomm does not signal to the participants that they consider policy X or Y important, they embolden editors to engage in improper behavior. Things can become worse during an Arbcomm proceeding because tensions are higher, and people want to "get even" with their perceived adversaries. And Arbcomm can act quickly if they want to, obviously. So why the long delays, even to produce some sort of innocuous general statements?